How simple is too simple?

I’m a big fan of the “release early, release often” philosophy, where you break your software into small chunks and release it to market as soon as you can, let people play with it and give you feedback, and iterate from there. The other common option is to build a huge release and push it all out there at once, and then only release changes every so often, and usually very large groups of changes.

I’ve been working on an idea with some friends for a few months now, which I’ll talk more about in a future post. Anyway, one of the constant challenges we face is where exactly to draw the line for our first release. We want to build something simple, but something that’s still useful. It can’t be too simple. As part of the process of deciding where to draw that line, I’ve been doing a lot of research and thinking a lot about other apps and sites and how they’ve dealt with the challenge of where to draw the line of simplicity. I wanted to give what I think is a good example of simple / too simple:


CheckYesOrNo is a small one-day project that was done by Ben Rasmusen and Gordon Brander, both friends of mine. It started out as a spur of the moment project that they quickly knocked out and put out there, which is awesome. It allows anyone to create a yes or no poll and solicit feedback from others via a URL. That’s it. There’s no registration, login, profiles, etc. Just a very simple concept and a beautiful design. However, once I used the app for a little while, I started to notice a small frustration: a lot of my questions didn’t fit into the “Yes or No” mold. Sometimes I had three alternatives that I wanted to ask people to choose between. Also, sometimes when I answered a poll, I wanted to provide more detail about my answer. And there was no way for me to keep track of the questions I had asked and the results other than keeping a list of URLs somewhere. In short, I felt like the app was just a little too simple for me.


Today, Ben released, which is still a very simple poll creator, with three main differences from CheckYesOrNo:

  • It allows you to define multiple answers (including yes or no, if you choose)
  • It allows those responding to the poll to leave short comments
  • It has a dashboard where you can track responses to all your polls

It still features a great design, clear functionality, and a solid user experience, but the usefulness is so much higher to me, just because it’s now just a bit more flexible, but virtually no usability has been sacrificed for the added functionality. It’s a great product and has drawn the line of simplicity quite skillfully.

In fact, I’ve been racking my brain for the last day or two and the only real enhancement I can think of at this point is to allow people to create polls without registering if they so desire. I’m sure other suggestions and improvements will emerge from people’s usage patterns and feedback, but that’s the whole point. When you throw it out there at the right point, it’s useful enough that people will actually use it and tell you what needs to be changed and improved. Any more simplistic and people won’t bother.

How simple is too simple to you?